The Trump administration removed export restrictions on Anthropic's advanced AI models Tuesday evening, restoring global customer access after a two-week ban that had created uncertainty across the technology sector. Commerce Secretary Howard Lutnick announced the decision in a letter to Anthropic, clearing the path for immediate deployment of Fable 5 and phased rollout of the more powerful Mythos 5 to select international partners.
The reversal reflects a deliberate shift in approach: rather than indefinite restrictions, the administration negotiated specific commitments with the company. Anthropic agreed to proactively detect and address security risks, coordinate with government on future releases, and report malicious activity. The company also implemented additional technical safeguards to block attempts to bypass Fable's cyber protections. Commerce's Center for AI Standards and Innovation tested both the previous and new safeguards before clearance.
The Path to Resolution
The export controls, issued June 12, had sparked immediate resistance from industry and allied governments. European nations particularly objected—they wanted access to Anthropic's tools to identify and patch vulnerabilities in their networks before adversaries could exploit them. The restrictions had created a chilling effect across the AI industry, raising questions about the administration's regulatory approach to cutting-edge technology development.
Lutnick emphasized the collaborative nature of the outcome. "We have worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the U.S. government and strengthen America's leadership in AI," he wrote. White House chief of staff Susie Wiles called it unprecedented private-sector cooperation, framing the resolution as foundational to an "America First" competitive strategy against China.
The administration had partially lifted restrictions on Mythos on June 28 before removing the Fable restrictions entirely. Negotiations had continued intensely over the weekend and into the week, indicating serious internal debate about balancing national security concerns with innovation incentives.
The Broader Regulatory Question
The resolution came amid broader administration moves to manage advanced AI deployment. OpenAI simultaneously agreed to restrict availability of its ChatGPT-5.6 model, rolling it out only to a limited group at White House request—a move CEO Sam Altman had resisted. The parallel actions suggest the administration is pursuing case-by-case negotiations rather than blanket policy.
Dean Ball, former Trump White House AI adviser and incoming head of strategic futures at OpenAI, expressed skepticism about the transparency of the Anthropic deal. "We have no idea what Anthropic did to make the models 'safe,' what commitments Anthropic has made going forward, and whether or how any of this applies to other frontier models in the government's licensing queue," he said.
Paul Lekas of the Software & Information Industry Association called the lifting of controls positive but warned of a larger problem: "There remained a real need for a consistent process and framework for frontier model assessment." Joe Hoefer, chief AI officer of Monument Advocacy, characterized the outcome as "more of a ceasefire than full resolution," suggesting ongoing uncertainty about long-term policy direction.
The Case for Market Competition
Glenn Gerstell, former general counsel of the National Security Agency and senior adviser at the Center for Strategic & International Studies, defended the administration's choice. "We're not going to maintain our lead over China in cutting-edge AI simply by slapping export controls on the latest innovation," he said. "The whole episode underscores how we have a lot more work to do to figure out the right way of balancing regulation for safety and national security reasons with promoting innovation in the AI sector."
This framing captures a core tension: heavy-handed export controls risk ceding global market share to competitors, particularly China, while inadequate safeguards create genuine security vulnerabilities. The Anthropic resolution attempts to thread that needle through negotiated commitments rather than blanket restrictions.
Why This Matters:
The decision signals the administration's preference for negotiated oversight over regulatory prohibition—a market-oriented approach that preserves American competitive advantage while addressing legitimate security concerns. However, the lack of transparent criteria for what constitutes "safe" AI deployment creates ongoing uncertainty for the industry. Companies face unclear expectations about future releases and unclear standards for government approval. The case-by-case approach, while flexible, prevents the kind of predictable regulatory environment that encourages long-term investment. Whether this model proves sustainable depends on whether the administration can establish consistent principles rather than ad-hoc negotiations. The stakes are significant: AI leadership is increasingly central to national competitiveness and security, and regulatory unpredictability could push innovation overseas or into less transparent channels.